A deep learning approach to predict significant wave height using long short-term memory

We present a framework for forecasting significant wave height on the Southwestern Atlantic Ocean using long short-term memory algorithm (LSTM), trained with ERA5 database available through Copernicus Climate Data Store (CDS) implemented by ECMWF (European Center Medium Range Forecast) and also buoy data. The predictions are made seven different locations in Brazilian coast, where data available, ranging from shallow to deep water. Experiments conducted exclusively historical series at selected influence of other variables as inputs training is investigated. results shows that data-driven methodology can be used surrogate computational expensive physical models, best accuracy near $95\%$, compared reanalysis